Deep Learning-Aided SCMA

Kim, Minhoe; Kim, Nam-I; Lee, Woongsup; Cho, Dong-Ho

doi:10.1109/lcomm.2018.2792019

Cited by 185 publications

(140 citation statements)

References 11 publications

Supporting

Mentioning

139

Contrasting

Unclassified

Order By: Relevance

“…We also note that DNNs that are proposed in communication systems could have many more parameters than the 9536 we use in this network. For example, [47] proposes an MLP network with 4 hidden layers where each layer has 512 neurons, which results in at least 512 × 512 × 3 = 786432 parameters. Thus a larger CNN can be trained to decode fixedlength capacity-approaching CS codes with longer codewords.…”

Section: Results and Outlookmentioning

confidence: 99%

Deep Learning-Based Decoding for Constrained Sequence Codes

Cao

Fair

2018

2018 IEEE Globecom Workshops (GC Wkshps)

View full text Add to dashboard Cite

Constrained sequence (CS) codes, including fixedlength CS codes and variable-length CS codes, have been widely used in modern wireless communication and data storage systems. Sequences encoded with constrained sequence codes satisfy constraints imposed by the physical channel to enable efficient and reliable transmission of coded symbols. In this paper, we propose using deep learning approaches to decode fixed-length and variable-length CS codes. Traditional encoding and decoding of fixed-length CS codes rely on look-up tables (LUTs), which is prone to errors that occur during transmission. We introduce fixed-length constrained sequence decoding based on multiple layer perception (MLP) networks and convolutional neural networks (CNNs), and demonstrate that we are able to achieve low bit error rates that are close to maximum a posteriori probability (MAP) decoding as well as improve the system throughput. Further, implementation of capacity-achieving fixedlength codes, where the complexity is prohibitively high with LUT decoding, becomes practical with deep learning-based decoding. We then consider CNN-aided decoding of variable-length CS codes. Different from conventional decoding where the received sequence is processed bit-by-bit, we propose using CNNs to perform one-shot batch-processing of variable-length CS codes such that an entire batch is decoded at once, which improves the system throughput. Moreover, since the CNNs can exploit global information with batch-processing instead of only making use of local information as in conventional bit-by-bit processing, the error rates can be reduced. We present simulation results that show excellent performance with both fixed-length and variable-length CS codes that are used in the frontiers of wireless communication systems.

show abstract

Section: Results and Outlookmentioning

confidence: 99%

Deep Learning-Based Decoding for Constrained Sequence Codes

Cao

Fair

2018

2018 IEEE Globecom Workshops (GC Wkshps)

View full text Add to dashboard Cite

show abstract

“…In addition, many specific aspects of communication systems are being studied from machine learning perspective, including modulation recognition [38], PAPR reduction [39], wireless interference identification [40], and so on. However, speaking to DL for SCMA, currently to the authors knowledge, very limited related works have been conducted or published besides the article [41], which only takes account of autoencoders without DCMA extending.…”

Section: Related Workmentioning

confidence: 99%

“…For this specific simulation, there are 512 hidden nodes for each hidden layer in D-SCMA+DNN, while DL-SCMA has only 48. There are 6 hidden layers each with 32 hidden nodes in the network in [41] for generating learned codebooks used by D-SCMA+MPA and D-SCMA+DNN, while the corresponding encoder part of AE-SCMA has 4 layers each also with 32 nodes. Therefore, in terms of computation, D-SCMA+DNN and its codebook generating DNN are more complex than the proposed DL-SCMA and AE-SCMA, respectively.…”

Section: Computational Complexitymentioning

confidence: 99%

A novel deep neural network based approach for sparse code multiple access

et al. 2020

View full text Add to dashboard Cite

Sparse code multiple access (SCMA) has been one of non-orthogonal multiple access (NOMA) schemes aiming to support high spectral efficiency and ubiquitous access requirements for 5G wireless communication networks. Conventional SCMA approaches are confronting remarkable challenges in designing low complexity high accuracy decoding algorithm and constructing optimum codebooks. Fortunately, the recent spotlighted deep learning technologies are of significant potentials in solving many communication engineering problems. Inspired by this, we explore approaches to improve SCMA performances with the help of deep learning methods. We propose and train a deep neural network (DNN) called DL-SCMA to learn to decode SCMA modulated signals corrupted by additive white Gaussian noise (AWGN). Putting encoding and decoding together, an autoencoder called AE-SCMA is established and trained to generate optimal SCMA codewords and reconstruct original bits. Furthermore, by manipulating the mapping vectors, an autoencoder is able to generalize SCMA, thus a dense code multiple access (DCMA) scheme is proposed. Simulations show that the DNN SCMA decoder significantly outperforms the conventional message passing algorithm (MPA) in terms of bit error rate (BER), symbol error rate (SER) and computational complexity, and AE-SCMA also demonstrates better performances via constructing better SCMA codebooks. The performance of deep learning aided DCMA is superior to the SCMA.As depicted in Fig. 1, consider J users transmitting data bits over the same K sub-carriers of OFDM, here K < J such that overloading is provided. According to SCMA encoder, each user maps every m = log 2 (M ) bits into a K-dimensional complex codeword c with only N non-zero elements standing for QAM modulation and LDS spreading combination, here N < K. The overlapping degree is d f = JN K , and overloading ratio is λ = J K . There are M codewords forming a codebook for each user and each codebook is unique. The encoding procedure can be described by c = f (b), where b ∈ B log2(M ) and c ∈ C ⊂ C K with |C| = M . Function f is actually a mapping matrix which

show abstract

“…The MMSE of the optimal one-step channel predictor for spatial multiplexing transmission at the k-th time index is given as (17). As seen, the Weiner filter predictor in (14) requires a priori knowledge about the channel statistics through matrixes A k−1 and R d k−1 .…”

Section: A Dd-ce For Spatial Multiplexing Using Wiener Predictormentioning

confidence: 99%

Decision Directed Channel Estimation Based on Deep Neural Network $k$ -Step Predictor for MIMO Communications in 5G

Mehrabi

Mohammadkarimi

Ardakani

et al. 2019

IEEE J. Select. Areas Commun.

View full text Add to dashboard Cite

We consider the use of deep neural network (DNN) to develop a decision-directed (DD)-channel estimation (CE) algorithm for multiple-input multiple-output (MIMO)-space-time block coded systems in highly dynamic vehicular environments. We propose the use of DNN for k-step channel prediction for space-time block code (STBC)s, and show that deep learning (DL)-based DD-CE can removes the need for Doppler spread estimation in fast time-varying quasi stationary channels, where the Doppler spread varies from one packet to another. Doppler spread estimation in this kind of vehicular channels is remarkably challenging and requires a large number of pilots and preambles, leading to lower power and spectral efficiency. We train two DNNs which learn real and imaginary parts of the MIMO fading channels over a wide range of Doppler spreads. We demonstrate that by those DNNs, DD-CE can be realized with only rough priori knowledge about Doppler spread range. For the proposed DD-CE algorithm, we also analytically derive the maximum likelihood (ML) decoding algorithm for STBC transmission. The proposed DL-based DD-CE is a promising solution for reliable communication over the vehicular MIMO fading channels without accurate mathematical models. This is because DNN can intelligently learn the statistics of the fading channels. Our simulation results show that the proposed DL-based DD-CE algorithm exhibits lower propagation error compared to existing DD-CE algorithms while the latters require perfect knowledge of the Doppler rate.part of the DD-CE is channel prediction. Existing channel predictors are highly depended on channel statistics which is severely affected by the estimation of Doppler spread of the channel. However, in highly dynamic vehicular environments, Doppler spread estimation is challenging.Recently, deep learning (DL) has been widely investigated in the signal processing and communications problems to improve the performance of some certain parts of conventional communication systems, such as decoding, estimation, and more [14]- [21]. In particular, DL-based CE methods have been studied in literature such as the recent work in [14]. A deep neural network (DNN) is a universal function approximator with superior logarithmic learning ability and convenient optimization capability, and thus can be used for the problems without any accurate mathematical model [22]. Currently, most of the existing algorithms in communications rely on precise mathematical models. However, in practice tractable mathematical models cannot reflect many imperfections and nonlinearities, and can only work as rough approximations when these issues are non-negligible. DL can fix this drawback in communication and information theory and offer algorithms without mathematically tractable models [18].Motivated by the limitations of existing channel predictors and the strength of DNN in learning and prediction, a DLbased DD-CE for MIMO STBC is proposed in this paper, where the MIMO channel coefficients are predicted by two trained DNNs. While existing c...

show abstract

Deep Learning-Aided SCMA

Cited by 185 publications

References 11 publications

Deep Learning-Based Decoding for Constrained Sequence Codes

Deep Learning-Based Decoding for Constrained Sequence Codes

A novel deep neural network based approach for sparse code multiple access

Decision Directed Channel Estimation Based on Deep Neural Network $k$ -Step Predictor for MIMO Communications in 5G

Contact Info

Product

Resources

About